by Greg Mayer
[Update: I’ve been informed by Sam Harris that Murray and Herrnstein did not make the quantitative genetics error I attributed to them: they did not suppose that in traits with high heritability mean differences between populations indicate genetic differences between populations. I wasn’t sure they did (not having read the book), hence my noncommittal “or at least their public proponents” caveat. I am happy to be corrected on this point. It makes the demonization of Sullivan even more perplexing.]
Sort of. We may read him, but we must find him “abhorrent”. Or at least so proclaims Ben Smith of the New York Times in “I’m Still Reading Andrew Sullivan. But I Can’t Defend Him.” The headline condemns Sullivan, in Smith’s voice no less, but then opens with Smith visiting Sullivan, apparently late at night, at Sullivan’s vacation home on Cape Cod. The whole piece contains these sorts of contradictions. Smith presents himself as a devotee of Sullivan, even an acolyte, and—maybe?—a friend. He goes on about how Sullivan pioneered political blogging, influenced a generation of writers, first made the case for marriage equality, presciently touted Obama’s importance before 2008, influenced votes in the Senate, and “helped lead America away from torture.”
But he is to be condemned. Why? Because in 1994, as editor of the New Republic, he published an excerpt from Charles Murray and Richard Herrnstein’s book The Bell Curve, along with some other articles discussing and critiquing the book.
I haven’t read the The Bell Curve, nor the New Republic from 1994 (and neither are high on my reading list). It’s well known that Murray and Herrnstein, or at least their public proponents, made some basic quantitative genetics errors, vitiating their main point on IQ and heritability (see details below). For having published this collection of articles, Sullivan is irredeemable.
Three years ago, when Murray (and his faculty interlocutor) were assaulted by students during a campus lecture gone bad at Middlebury College, Sullivan commented on the 1994 publication, and this, I suppose, is his current view (Sullivan rarely writes about Murray):
[P]rotests against Murray are completely legitimate. The book he co-authored with Harvard professor Richard Herrnstein more than 20 years ago, The Bell Curve, included a chapter on empirical data showing variations in the largely overlapping bell curves of IQ scores between racial groups. Their provocation was to assign these differences to both the environment and genetics. The genetic aspect could be and was exploited by racists and bigots.
I don’t think that chapter was necessary for the book’s arguments, but I do believe in the right of good-faith scholars to publish data — as well as the right of others to object, critique, and debunk. If the protesters at Middlebury had protested and disrupted the event for a period of time, and then let it continue, I’d be highly sympathetic, even though race and IQ were not the subject of Murray’s talk. If they’d challenged the data or the arguments of the book, I’d be delighted.
Whether this reflects Sullivan’s view in 1994, I’m not sure. But I do know that Sullivan is a Voltairean proponent of free speech, and would publish things he disagreed with. Indeed, Sullivan’s willingness to entertain opposing views, and then to change his mind, is one of the striking things about him. It is thus curious that Smith, who claims that Sullivan was an “obvious influence” on his own work, misses this key element of Sullivan’s writing. Oddly, Smith goes on about how Sullivan never changes; but one of the most striking admissions of error I have ever seen in the public intellectual sphere is Sullivan’s owning up to the failures—not just in execution, but in conception—of the neoconservative project for Iraq that he had once so loudly cheered for.
Smith also writes about Sullivan’s departure from New York magazine, which Smith portrays as a preemptive firing before the “woke” staff at the magazine could organize against him. Sullivan’s publication of the New Republic issue—in 1994—was “a firing offense”— in 2020. If true, this is amazing, and damning of New York, its staff, and its editors. (Smith’s account is based on two anonymous “senior employees”.)
Smith’s piece is very Times à la 2020: you’re intolerable because I don’t like something you did more than 25 years ago. And in this case, it’s uber-Times à la 2020—it’s not anything you said 25 years ago, but the fact that you let someone else say something that I don’t like 25 years ago. At the Times, editors must pay for the opinions of authors!
Smith at times seems conflicted in carrying out his hatchet job on Sullivan—having praised him, even defended him, he seems reluctant to bring the dagger down, but ultimately he does. Smith is, perhaps, afraid for his own job. The Times has shown itself to be all too willing to force out or fire those who do not toe the line—just ask James Bennett.
I had read some of Ben Smith’s writing before he moved to the Times last spring, and I thought that this move would be a good thing. But it may be that rather than Smith being good for the Times, the Times has been bad for Smith.
Very brief quantitative genetics lesson: The basic error is to mistake heritability, a within population measure of the proportion of phenotypic variation in the population attributable to genetic differences, for a measure of the degree of genetic differentiation among populations; it is not. As a simple example, suppose we had two populations of a grass that have the same distribution of genotypes, and that heritability of height is high (and furthermore, to be absolutely explicit, there is no genotype X environment interaction). One population is grown with fertilizerm and the other without fertilizer. Although height has high heritability in both populations, the difference in mean height between populations is entirely environmental.